#4449: GHC 7 can't do IO when demonized
----------------------------+-----------------------------------------------
Reporter: kazu-yamamoto | Owner: adept
Type: bug | Status: new
Priority: normal | Milestone: 7.0.2
Component: Compiler | Version: 7.1
Resolution: | Keywords: forkProcess
Testcase: | Blockedby:
Difficulty: | Os: MacOS X
Blocking: | Architecture: x86
Failure: None/Unknown |
----------------------------+-----------------------------------------------
Comment(by adept):
More info: the first two calls for `closeControl` are made from `withNew`
(System/Event/Manager.hs). I don't know enough about weak IORefs to
understand what's happening to them over fork, but this is definitely a
root cause. When innermost process tries to initialize its event manager,
the following code is executed:
{{{
newWith be =
...
_ <- mkWeakIORef state $ do
st <- atomicModifyIORef state $ \s -> (Finished, s)
when (st /= Finished) $ do
I.delete be
closeControl ctrl
...
}}}
and after that something (my guess is `threadDelay`) forces the value
inside that weak IO Ref, pulling state corresponding to the "parent" io
managers through `I.delete` and `closeControl`.
Since two "outer" processes have the same kqueue FD numbers, repeated
`I.delete` produces the error we are observing.
So far, I am at a loss how to fix this.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/4449#comment:33>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
